Abstract 185P
Background
This modeling study aimed to assess the clinical and cost consequences of a hypothetical AI-assisted image reading solution in Netherlands-based lung cancer screening (LCS), comparing it to traditional image reading without AI. While low-dose computed tomography (LDCT) in LCS has shown a 20-24% reduction in lung cancer mortality, it can put a strain on radiologists' workload. Despite promising results, AI-assisted image reading is not widely integrated into clinical practice.
Methods
We conducted a cost-consequence analysis, considering costs and effects of different LCS scenarios from a healthcare perspective. Key model inputs included the eligible and screening populations, radiologists' image reading time, average weighted time, AI's image reading time, costs, screening effectiveness without AI, and discrepancies in image reading. The control scenario involved LCS without AI, while Scenario A introduced AI as a parallel reader. In Scenario B, AI served as the first reader, with a radiologist confirming positive scans and identifying false-positive classifications.
Results
LCS with AI-assisted image reading demonstrated potential cost reductions of 37% and 73% in Scenario A and B, respectively (total reading costs [Control: €29,676,879; Scenario A: €18,704,843; Scenario B: €8,146,251]). Additionally, integrating AI as the first reader could alleviate the workload on radiologists.
Conclusions
The integration of AI-assisted image reading in LCS presents significant cost reductions associated with image reading, supporting the use of AI to ease constraints on healthcare resources.
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.